DESCRIPTION: (Applicant's Abstract) Medulloblastoma, the most common malignant central nervous system neoplasm in children, has a dismal prognosis despite multi-institutional clinical trials using increasingly sophisticated neurosurgical and therapeutic interventions, with the majority of children with advanced medulloblastomas eventually dying of progressive disease. Progress has been hindered by technical inabilities to discriminate prospectively between patients with aggressive and indolent neoplasms. Recent advances in molecular biology, cytogenetics, and immunohistochemistry now convincingly demonstrate the phenotypic and genotypic heterogeneity of these neoplasms. Cytogenetic analysis and restriction fragment length polymorphism analysis using Microsatellite probes have shown that approximately 1/3 of these tumors have loss of the 17p chromosomal arm. Approximately 5% of medulloblastomas have mutations of the p53 gene, which is located on 17p and approximately 5% of these tumors have amplification of myc family genes. Preliminary studies from the applicant's laboratory and others suggest that loss of heterozygosity (LOH) for 17p, may identify patients with aggressive neoplasms and significantly shorter survivals. Furthermore, recent immunohistochemical phenotypic studies suggest that evidence of glial or neuroblastic differentiation in the desmoplastic pattern may predict a long-term survival. The applicant's hypothesis is that patients with medulloblastomas exhibiting LOH for 17p, have a significantly worse prognosis than patients whose tumors lack these alterations and that the identification of neuroblastic or glial differentiation by immunohistochemistry can be used to define these prognostic groups more fully. In Specific Aim 1 she will analyze paraffin-embedded human medulloblastoma specimens for LOH of chromosome 17p using microsatellite markers, fluorescence in situ hybridization (FISH) and comparative genomic hybridization (CGH). In Aim 2, she will reclassify all medulloblastomas as "classic," "desmoplastic", "mixed", or "other" and will define, by immunohistochemistry, the distribution of synaptophysin, neuron-specific enolase, neurofilament proteins, and glial fibrillary acidic protein, in these specimens. Using these data, in Specific Aim 3, the applicant will define the biological relevance of LOH for 17p, and phenotypic profile, with regard to initial tumor staging, tumor recurrence, and clinical outcome. She believes that these studies will provide a rational, molecular-based classification scheme for these neoplasms which has prognostic significance. In addition, these observations will establish the scientific basis for further investigations into the basic mechanisms of medulloblastoma etiology and progression.